How to measure ambient noise level

[Niko Busch]

We are currently setting up our new Biosemi equipment in a new testing room. I would like perform a "phantom" recording - no human subject - to get a feeling of the ambient noise level of the room. Thus, I am mostly interested in 50Hz noise, monitor noise and 16.7 tram line noise. I already tried the one-bucket-test as specified in the manual, using the CMS/DRL electrodes and 3 flat active electrodes. After recording, I first referenced the signals in EEGLAB and then computed FFT spectra. To my surprise, the spectra we basically flat - not even the smallest peak at 50 Hz. This was even when I put several electrical devices right next to the bucket (monitor, hair dyer, etc.). I find this surprising, because with our previous setup, we always had a very small and tolerable but still clearly visible 50 Hz peak in the spectra, when we recorded real subjects.

Does this result indicate that using a bucket of salt walter, Biosemi perfectly reduces all ambient noise, or is the one-bucket-test simply not a solution for measuring ambient noise?

[Coen]

A container with salt water is a good electrical approximation of the human body. However, interference levels depend on capacitive couplings of the human body with the inference sources. For an improved test, your water bucket should approximate the size, shape and position of the human body.

Best regards, Coen (BioSemi)